Pré-Publication, Document De Travail Année : 2026

Combining machine learning quantile regression and Gaussian random fields: a general framework for modeling and simulating space-time processes

Résumé

Modeling and simulating space-time random fields while accounting for possibly misaligned covariates is crucial for many applications requiring uncertainty quantification and risk assessment. To achieve this, we propose a flexible framework that couples machine-learning quantile regression with space-time Gaussian random fields.

In this framework, the target variable is modeled as a combination of a latent Gaussian random field and transformed marginals obtained by Machine learning quantile regression conditionally on a set of covariates. We illustrate the approach on a synthetic experiment and on a case study that considers daily maximum temperature over north-east Switzerland conditional on seasonal cycle and large-scale geopotential height.

Fichier principal
Vignette du fichier
Quantile Gneiting.pdf (26.59 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05441043 , version 1 (05-01-2026)

Licence

Identifiants

  • HAL Id : hal-05441043 , version 1

Citer

Said Obakrim, Denis Allard, Lionel Benoit, Gregoire Mariethoz. Combining machine learning quantile regression and Gaussian random fields: a general framework for modeling and simulating space-time processes. 2026. ⟨hal-05441043⟩
318 Consultations
36 Téléchargements

Partager

  • More